SURFACE SEISMIC AND ELECTRICAL METHODS TO DETECT FLUIDS RELATED TO FAULTING

In the absence of drilling, surface-based geophysical methods are nece ssary to observe fault zones and fault zone physical properties at sei smogenic depths. These in situ physical properties can then be used to infer the presence and distribution of fluids along faults, although such observations are by nature indirect and become less exact with gr eater depth. Multiple observations of a range of such geophysical prop erties as compressional and shear seismic velocity (V-p and V-s), V-p/ V-s ratio (related to Poisson's ratio), resistivity and attenuation in and adjacent to fault zones offer the greatest hope of making inferen ces of the fault zone geometry, fluids in the fault zone, and fluid re servoirs in the surrounding crust, For simple geometries, fault zone g uided waves can provide information on fault zone width and velocities for faults of the order of 200 m wide. To address the question of whe ther a narrow fault zone can be imaged well enough at depths of seismi c rupture to infer the presence of anomalously high fluid/rock ratios, we present synthetic seismic tomography and magnetotelluric examples for an ideal case of a narrow fault zone with a simple geometry, large changes in material properties, and numerous earthquakes within the f ault zone. A synthetic 0.5-km wide fault zone with 20% velocity reduct ion is well imaged using local earthquake tomography. When sequential velocity inversions are done, the true fault width is found, even to 9 km depth, although the calculated amplitude of the velocity reduction is lower than the actual amplitude. V-p/V-s is as well determined as V-p Magnetotelluric imaging of a synthetic fault zone shows that a con ductive fault zone can bk well imaged within the upper 10 km. Further, a narrow (1 km) very low resistivity (3 ohm m) fault core can be imag ed within a broad (5 km) low resistivity (10 ohm m) fault zone, illust rating that regions of a fault containing large quantities of intercon nected fluids within a broader, conductive fault zone should be detect able. Thus variations in fluid content and fluid pressure can be infer red from electrical and seismic methods but there will always be uncer tainty in these inferences due to the trade-off with other factors, su ch as intrinsic variations in porosity, mineralogy, and pore geometry. The best approach is combined modeling of varied seismic and electric al data.